I was talking with a friend earlier, discussing the effect of the position of the center of gravity on the takeoff distance of an airplane, and whether forward cg or aft cg would result in the shortest ground roll during takeoff. I first deduced that aft cg would result in the lowest tail loads during rotation, therefore producing the least drag and the shortest ground roll in the same way that this situation is favorable for cruise. If this is the case, then forward cg would probably be better for landing?

Then it occurred to me that on a softer surface in a tail-wheel airplane, an aft cg may produce more rolling resistance on the tail-wheel due to extra down force causing a deeper depression on the earth beneath it, as opposed to a greater proportion of the weight on the usually much larger mains. Following this logic, aft cg in a tricycle would relieve the load on the nose wheel, reducing the tendency to wheel-barrow through the soft stuff.

I have not seen any studies on this topic, and perhaps the effect is totally negligible, but I was curious if any BCP dwellers might possess this wisdom, have conducted tests, or has thoughts on this matter?

Aft CG results in the lowest stall speed and thus, if flying at that lower speed, would result in the shorter T/O and landing. Depending on the craft I suppose it might feel less stable and one might not feel as comfortable at the slower speed? The FAAs Pilot's Handbook of Aeronautical Knowledge probably has some nice diagrams to remind how the wing has to lift the weight of the craft PLUS whatever negative lift created by the tail in a conventional design aircraft (non-canard). So just like you notice the effect on T/O distance when you fill the tanks or when your buddy comes along, a forward CG adds to the wing loading when compared with a rearward CG.

Pilot's Handbook of Aeronautical Knowledge
Figure 4-55 illustrates why this is true. With forward loading, “nose-up” trim is required in most aircraft to maintain level cruising flight. Nose-up trim involves setting the tail surfaces to produce a greater down load on the aft portion of the fuselage, which adds to the wing loading and the total lift required from the wing if altitude is to be maintained. This requires a higher AOA of the wing, which results in more drag and, in turn, produces a higher stalling speed.
With aft loading and “nose-down” trim, the tail surfaces exert less down load, relieving the wing of that much wing loading and lift required to maintain altitude. The required AOA of the wing is less, so the drag is less

Actually, as you've already pointed out....It Depends.

MTV

I've noticed I use pretty much the same length of runway (when landing) when carrying a passenger. Since we are talking about a 719 lb airplane and 200' landings that struck me as significant. I figured it had something to do with the center of gravity, expecially since my S-7S is a taildragger, The trim drag theory, or lack thereof, that Matt quotes sure sounds right to me. Its always good to have a gut feeling officially verified!